THE NEOTROPICAL REGION 81 



same Order, which seem to be more nearly related to the 

 Australian marsupials of other families ; in fact, Ame- 

 ghino has placed some of them in the (hitherto supposed 

 to be) exclusively Australian family Dasyuridte. 



The fossil Edentates of this formation are all of the 

 American section of the group. The Perissodactyle 

 Ungulates are represented by two families (Protero- 

 theriidx and Macraucheniidi&)riot very definitely connected 

 with other Perissodactyles, and unknown elsewhere ; and 

 the Toxodonts (another group forming a distinct sub-order 

 of the Ungulata) are also quite unrecognised outside 

 South America, where, indeed, they appear to have existed 

 for a short period only. 



The Rodents of this formation all belong to the Hystri- 

 comorphine section of the order, and the Monkeys are 

 all Platyrrhine, so that, except in the case of the Marsupials 

 where the results are to a certain extent doubtful, no clue 

 to the origin of the Neotropical fauna is shown by this 

 extinct fauna. When these results are compared with 

 the Eocene Mammal -fauna of the northern hemisphere, 

 the absence of Artiodactyles, Insectivores, Bats, Carnivores, 

 and Lemurs, is very striking. 



The " Patagonian " beds contain a Mammal-fauna only 

 to be distinguished from the previous " Santa Cruz " series 

 by a further differentiation of genera belonging to the same 

 orders without any traces of foreign admixture. When, 

 however, we reach the more recent Araucanian formation, 

 we find, in addition to the Edentates, Toxodonts, and 

 other typical South American forms, a number of foreign 

 intruders, such as Tapirs, Lamas, Elephants {Mastodon), 

 and Wild Dogs (Canis), of an entirely different aspect. 

 There can be little doubt that these animals had migrated 



F 



